Researchers at the University of Copenhagen and University of Colorado Boulder have identified a temperature-independent source of energy loss limiting the performance of hybrid superconductor-semiconductor qubits, known as gatemons. The work demonstrates that gatemons, designed for improved control, consistently exhibit significantly shorter coherence times, in the few-microsecond range, compared to traditional transmon qubits which reach tens of microseconds in the same architecture.

This discrepancy arises from dissipation within the junction itself, exceeding losses from other sources by more than ten-fold. The findings indicate that simply cooling these qubits will not resolve the issue, as the dissipation remains constant regardless of temperature.

Hybrid Gatemon and Transmon Co-Fabrication for Direct Comparison

Co-fabrication of gatemon and transmon qubits with identical layouts revealed a significant disparity in coherence times, with transmons achieving relaxation times in the tens of microseconds while gatemon saturation occurred in the few-microsecond range. This direct comparison, detailed in a recent study, was designed to isolate the impact of the Josephson junction material, the key distinction between the two qubit types, on overall performance. The research team constructed a loss budget encompassing Purcell decay, spontaneous emission through control lines, and internal dielectric loss to understand the observed differences.

The study’s methodology allowed researchers to use the transmons as on-chip references, providing a direct means of assessing the limitations of the gatemon design. Temperature-dependent measurements of $T_{1}$ , a measure of qubit relaxation time, followed a quasiparticle-activation model for both gatemon and transmon qubits, yielding similar superconducting gaps for both S-Sm-S and SIS junctions.

This similarity rules out a substantially reduced induced gap as the primary cause of the shorter gatemon lifetimes, focusing attention on other potential loss mechanisms. The persistent discrepancy points to junction-intrinsic dissipation as a critical factor limiting gatemon performance. Researchers write that relaxation limitations in gatemons may arise from junction-intrinsic mechanisms such as finite subgap conductance and imperfect superconductor-semiconductor interfaces, suggesting that these imperfections create a residual density of states and low-energy excitations.

This finding challenges the assumption that simply lowering the operating temperature will resolve coherence issues in gatemon qubits, as the dissipation remains largely unaffected by temperature changes. Improving coherence in hybrid qubits will likely require a focus on engineering cleaner interfaces between materials, suppressing subgap conductance and implementing strategies to manage quasiparticles.

The co-integrated gatemon-transmon platform developed in this work offers a valuable tool for quantitatively benchmarking junction-intrinsic dissipation in future hybrid superconducting qubits. The researchers believe this platform provides a blueprint for future investigations aimed at optimizing the performance of these promising qubit designs.

Gate Tunability via S-Sm-S Junction Energy-Phase Relation

Hybrid superconductor-semiconductor junctions enable in-situ gate tunability of the Josephson energy, a feature exploited in the creation of gatemons, Andreev spin qubits and fluxonium qubits. The energy-phase relation within these junctions, defined by parameters like transmission probabilities and gate voltage, dictates their behavior, but a persistent discrepancy between gatemon and transmon qubit lifetimes has remained unexplained. The fabrication of gatemons and transmon qubits on identical silicon chips, sharing circuit designs, fabrication steps and gate dielectric layers, allowed for a direct comparison with the Josephson junction as the sole intentional distinction between the two.

This co-integrated platform facilitated systematic analysis of relaxation times, $T_{1}$ , across multiple devices and material stacks. Transmon qubits in this architecture achieve relaxation times in the millisecond regime, a result of mature junction fabrication, controlled dielectric participation and mitigation of known loss channels.

In contrast, gatemons consistently saturate in the few-microsecond range, prompting a detailed loss budget analysis. The resulting $T_{1}$ limits for the transmons exceeded measured gatemon values by more than an order of magnitude, indicating a fundamental difference in dissipation mechanisms. The team redesigned the gatemon gate layout to reconcile strong DC tunability with suppressed spontaneous emission, implementing an XYZ-style control-line architecture optimized for both qubit types.

However, the data strongly suggests that reduced coherence in gatemons stems from temperature-independent, junction-intrinsic dissipation. The close proximity of the gate electrode to the junction, necessary for efficient modulation of carrier density in the semiconductor, presents a challenge for gatemons.

On-Chip Loss Budget Reveals Discrepancy in T_1 Limits

This comparison, conducted across multiple chips with identical circuit layouts, gate dielectrics and control lines, isolated the Josephson junction as the primary differentiating factor between the two qubit designs. This finding prompted a deeper investigation into junction-specific dissipation as a potential limiting factor, leading to the inclusion of a single, frequency-independent loss term associated with the superconductor-semiconductor interface. By incorporating this additional loss channel, the model accurately reproduced measured gatemon $T_{1}$ values, suggesting a dominant role for junction-intrinsic dissipation in limiting qubit coherence.

Further analysis revealed that this dissipation is not mitigated by lowering temperatures, indicating a temperature-independent mechanism at play. The study reports that by augmenting this budget with a single phenomenological, frequency-independent loss term associated with the S-Sm-S junction, they obtained a curve that reproduces a representative gatemon $T_{1}$ , emphasizing the persistence of the loss even at cryogenic temperatures. The researchers used optical imaging to confirm the uniformity of the fabricated samples, ensuring that variations in fabrication processes did not contribute to the observed differences.

Temperature-Dependent T_1 and Quasiparticle Activation

Temperature-dependent measurements of qubit relaxation times reveal a stark contrast between hybrid superconductor-semiconductor gatemons and conventional transmons, with gatemons exhibiting significantly shorter lifetimes even at cryogenic temperatures. Specifically, transmon qubits fabricated alongside gatemons on Chip 2 achieved relaxation times of approximately 20 microseconds at base temperature, while the gatemons plateaued at just a few microseconds, around 3.70 and 4.04 microseconds for the high- and low-frequency configurations, respectively.

This discrepancy is not simply a matter of temperature, as the observed gatemon $T_{1}$ values remain substantially lower than predicted by models accounting for Purcell decay, spontaneous emission and internal dielectric loss. The research team investigated whether quasiparticle poisoning, a known limitation in superconducting qubits, could explain the reduced coherence in gatemons, and performed temperature-dependent $T_{1}$ measurements on both gatemon and transmon devices co-fabricated on the same chip.

Both qubit types initially demonstrated a temperature-independent $T_{1}$ plateau at low temperatures, but exhibited a pronounced decrease as temperatures increased, consistent with thermally activated quasiparticles. The team modeled this behavior using a quasiparticle-activation model, finding similar superconducting gaps for both the S-Sm-S junctions in the gatemons and the SIS junctions in the transmons, indicating that the reduced gatemon coherence stems from additional dissipation channels.

Calculations based on this finding indicate an upper bound of approximately 10^-5 for the nonequilibrium quasiparticle population, a value several orders of magnitude higher than typically observed in standard transmons. The researchers report that at base temperature, the transmon exhibits, whereas the gatemon saturates at and for the two operating points, captured by a significantly larger $Γ_{0}$ for the gatemon, highlighting the magnitude of the difference.

This finding is critical because it points to a fundamental materials limitation within the gatemon design, a limitation not as prominent in traditional transmon qubits, and suggests that future improvements will require a deeper understanding of the semiconductor-superconductor interface. The observed temperature dependence, while consistent with quasiparticle activation, ultimately emphasises the dominance of this junction-specific dissipation in limiting the performance of these hybrid qubits.

Junction-Intrinsic Dissipation Dominates Gatemon Coherence

Measurements of paired gatemon and SIS-junction transmon qubits reveal a temperature-independent dissipation mechanism limiting coherence in the hybrid devices, despite similar superconducting gaps between the two junction types. The work demonstrates that, while both qubit designs initially exhibit a temperature-independent T₁ plateau at low temperatures, the gatemon’s relaxation times plateau at significantly shorter durations than those achieved by the transmon qubits. This discrepancy, observed across multiple chips, indicates that the source of coherence loss is not simply a reduced superconducting gap in the gatemon’s semiconductor-superconductor junction.

The team’s calculations show that the Purcell-limited relaxation time, T_1,Pur, exceeds a value of 10 microseconds over a frequency range of 3 to 5 GHz, as indicated by dotted lines in their detailed gatemon qubit design. The team considered spontaneous emission through the control line, calculating the associated relaxation time as T_1,spont =, where C_q represents the total qubit capacitance and Y is the admittance to the control line as seen by the qubit.

The remaining difference in coherence times, after accounting for Purcell decay and spontaneous emission, points to an additional temperature-independent decay rate specific to the gatemon qubits. This finding is critical for advancing hybrid qubit technology, as it identifies a fundamental materials limitation that must be addressed to achieve coherence times comparable to those of conventional transmon qubits.

Purcell Decay and Dielectric Loss Fail to Explain T_1 Difference

The discrepancy in coherence between gatemons and transmons extends beyond expected losses from Purcell decay, spontaneous emission and dielectric dissipation, according to new analysis. This systematic offset suggests that the limiting factor in gatemon performance is an intrinsic property of the superconducting-semiconductor junction itself, not a broadly applicable environmental effect. A detailed loss budget, constructed using the on-chip transmon performance as a benchmark, revealed that established loss mechanisms fall short of explaining the observed difference in T₁.

The team’s analysis considered the energy decay rate, expressed as Q = 2π = 2π, and explored how independent loss channels contribute to the overall quality factor. Further investigation focused on temperature-dependent measurements, revealing a consistent pattern of dissipation that remains largely unaffected by cooling. The researchers found that thermally activated quasiparticles, while present, do not fully explain the reduced gatemon coherence, pointing to an additional, temperature-independent loss channel.

This suggests that junction-intrinsic dissipation, captured by a phenomenological parameter, is the dominant source of coherence loss in gatemons. The team emphasizes that the precise value of this additional loss, termed Q_phenom, is device-dependent, but its presence is sufficient to reconcile the data with an otherwise standard loss budget.

They note that while the microscopic mechanism behind Q_phenom remains unclear, its impact is significant, and is further constrained by temperature-dependent measurements. The findings highlight a fundamental trade-off in hybrid qubit design, where the pursuit of improved gate tunability through superconductor-semiconductor junctions introduces a new and substantial limitation on qubit lifetimes, a challenge not as prevalent in traditional transmon qubits.

👉 More information
🗞 Junction-intrinsic dissipation in hybrid superconductor-semiconductor gatemon qubits
✍️ Zhenhai Sun et al.
🧠 DOI: http://link.aps.org/doi/10.1103/dmnb-x3km

More like this